Abstract
BACKGROUND: Whether ascites volume independently predicts survival in ovarian cancer remains unresolved due to conflicting evidence, binary classification approaches, and the absence of dose-response analysis. This prospective study aimed to characterize the dose-response relationship between ascites volume and survival outcomes and identify clinically meaningful volume thresholds. METHODS: We analyzed data from 293 high-grade serous ovarian cancer patients from Xiangya Hospital, Central South University [2017-2020]. Patients were stratified by intraoperative ascites volume: no ascites (0 mL), low-volume (<1,000 mL), and high-volume (≥1,000 mL). The primary outcomes were progression-free survival (PFS) and overall survival (OS). Restricted cubic spline (RCS) analysis examined dose-response relationships between ascites volume and survival outcomes. Cox proportional hazards models estimated survival associations. RESULTS: Among 293 patients, 15.4% had no ascites, 56.0% had low-volume, and 28.7% had high-volume ascites. RCS analysis revealed significant nonlinearity for PFS, with hazard ratio (HR) increasing steeply from 0 to 1,000 mL then plateauing (P for nonlinear =0.004), while OS demonstrated a predominantly linear relationship (P for nonlinear =0.83). Multivariable Cox analysis indicated that high-volume ascites independently predicted worse outcomes (adjusted HR 1.29 for PFS, 95% confidence interval (CI): 1.02-1.64, P<0.05; adjusted HR 1.54 for OS, 95% CI: 1.07-2.23, P<0.05), whereas low-volume ascites showed no independent prognostic significance. Adjusted 3-year PFS rates were 44.2%, 40.8%, and 36.5% for the three groups, respectively, while adjusted 3-year OS rates were 83.6%, 81.9%, and 76.5%. CONCLUSIONS: Ascites volume represents a quantifiable, independent prognostic factor in ovarian cancer with distinct dose-response patterns: a nonlinear relationship with PFS characterized by a critical 1,000 mL threshold, and a continuous linear relationship with OS.